Article
Computer Science, Artificial Intelligence
Wenxin Wang, Huachao Dong, Peng Wang, Jiangtao Shen
Summary: This paper proposes a bi-indicator-based surrogate-assisted evolutionary algorithm (BISAEA) for solving computationally expensive multi-objective optimization problems (MOPs). BISAEA utilizes a Pareto-based bi-indicator strategy and a radius-based function (RBF) model to approximate objective values. It also incorporates a one-by-one selection strategy based on angles and Pareto dominance to improve diversity. Experimental results show that BISAEA achieves high efficiency and a good balance between convergence and diversity. Application of BISAEA to a multidisciplinary optimization problem further demonstrates its superior performance on computationally expensive engineering problems.
COMPLEX & INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Interdisciplinary Applications
Kuangnan Fang, Xinyan Fan, Shuangge Ma, Qingzhao Zhang
Summary: In this study, a robust statistical model is adopted to handle longitudinal data, taking into account the long-tailed/contaminated distributions and interconnections among covariates. The model uses a network structure and novel penalties to model the interconnections, and numerical studies demonstrate its competitive practical performance.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2022)
Article
Computer Science, Interdisciplinary Applications
Daeju Kim, Shuichi Kawano, Yoshiyuki Ninomiya
Summary: A basis expansion with regularization methods is proposed for flexible or robust nonlinear regression models. When the underlying function has inhomogeneous smoothness, conventional regularization methods tend to over-fit or underfit certain parts. Therefore, a smoothly varying regularization method with adaptive penalties is considered. The proposed method is shown to perform well in various situations through simulations and real data analysis.
COMPUTATIONAL STATISTICS & DATA ANALYSIS
(2023)
Article
Computer Science, Hardware & Architecture
Monday Eze, Charles Okunbor, Deborah Aleburu, Olubukola Adekola, Ibrahim Ramon, Nneka Richard-Nnabu, Oghenetega Avwokuruaye, Abisola Olayiwola, Rume Yoro, Esomu Solomon
Summary: This work demonstrates the practical applications and evolutionary concepts of Parametric Curves, specifically focusing on higher order parametric Bezier curves. The relevance of Bezier curves in various fields such as computer graphics, robotics, animations, and virtual reality was explored. The work also addresses evolutionary issues related to parametric equations, theorem proofs, and calculus computations. A practical case study showcasing the implementation of quadratic Bezier curves in graphical design and programmatically evolving handless cups is presented.
COMPUTER SYSTEMS SCIENCE AND ENGINEERING
(2022)
Article
Mathematics
Jiajia Zhang, Yuanhua Qiao, Lijuan Duan, Jun Miao
Summary: The study focuses on bifurcation diagrams and exact multiplicity of positive solutions for the one-dimensional prescribed mean curvature equation with different parameters, revealing the shape characteristics of the bifurcation curve under various conditions.
Article
Mathematics
Zhaohao Wu, Deyun Zhong, Zhaopeng Li, Liguan Wang, Lin Bi
Summary: This paper presents a method for orebody implicit modeling based on normal estimation of cross-contour polylines. The method can automatically estimate normals and reorient them, showing advantages of low calculation, high efficiency, and strong reliability.
Article
Remote Sensing
Munmun Baisantry, Anil K. Sao, Dericks Praise Shukla
Summary: In this work, a band selection technique based on FDA and functional PCA is proposed. The method selects shape-preserving, discriminative bands which can highlight the important characteristics, variations, and patterns of the hyperspectral data, resulting in improved classification accuracy.
INTERNATIONAL JOURNAL OF REMOTE SENSING
(2022)
Article
Computer Science, Artificial Intelligence
Pengfei Huang, Handing Wang, Yaochu Jin
Summary: This study introduces semi-supervised learning to offline data-driven evolutionary optimization by utilizing the tri-training algorithm to overcome data deficiency. Experimental results demonstrate that the proposed method is competitive on problems with up to 500 decision variables.
SWARM AND EVOLUTIONARY COMPUTATION
(2021)
Article
Computer Science, Information Systems
Zan Yang, Haobo Qiu, Liang Gao, Danyang Xu, Yuanhao Liu
Summary: This paper proposes a general framework of surrogate-assisted evolutionary algorithms (GF-SAEAs) to adaptively arrange search strategies based on actual simulation cost differences. It classifies all constraints and designs a level-by-level feasible region-driven local search strategy to locate potential sub-feasible regions. Three different search mechanisms are employed to explore and exploit these located regions. Experimental studies show that GF-SAEAs outperform other state-of-the-art algorithms.
INFORMATION SCIENCES
(2023)
Article
Ornithology
Heloisa Helena Linhares, Esteban Frere, Ana Milliones, Gisele Pires de Mendonca Dantas
Summary: Kelp Gull is a widely distributed gull species in the Southern Hemisphere. The number of its subspecies is still controversial. This study aimed to evaluate the subspecies of Kelp Gull using mtDNA and investigate its demographic history and population structure in the Southern Hemisphere. Through sequencing Cytochrome b in 98 samples from Brazil, Argentina, and Antarctica, and adding 20 haplotypes from GenBank, the study found no support for the proposed subspecies clades based on Bayesian Phylogeny. However, genetic population structure of Kelp Gull in the Southern Hemisphere can be observed based on haplotype frequency. Additionally, there is evidence of genetic diversity loss in Kelp Gull during the Holocene population expansion around 2500-3000 years ago.
JOURNAL OF ORNITHOLOGY
(2023)
Article
Energy & Fuels
Guodong Chen, Xin Luo, Jiu Jimmy Jiao, Xiaoming Xue
Summary: The study proposes a GDDE algorithm to reduce the number of simulation runs in well-placement and control optimization problems, utilizing PNN as a classifier to select candidates, and building a local surrogate model with RBF to accelerate convergence. The algorithm shows a promising perspective in reducing simulation runs and improving optimization efficiency.
Article
Energy & Fuels
Emre Akarslan, Rasim Dogan
Summary: Identification of load in a grid is crucial for grid management and security. A novel load identification model combining feature selection method and classifier achieves over 95% identification accuracy. Experimental results show that using the combined model can yield better results with less data, and incorporating feature selection method in the classification schema improves load identification performance.
SUSTAINABLE ENERGY GRIDS & NETWORKS
(2021)
Article
Computer Science, Artificial Intelligence
Jinglu Li, Peng Wang, Huachao Dong, Jiangtao Shen
Summary: This paper proposes a multi/many-objective optimization algorithm assisted by radial basis function based on reference vectors to solve computationally expensive optimization. The algorithm determines a set of candidates guided by reference vectors in a sub-cycle. It utilizes a refinement regeneration strategy and a dynamic exploration strategy based on the candidate pool. The final Pareto-optimal solutions are obtained by repeatedly selecting candidates and applying the refinement regeneration and dynamic exploration strategies. The proposed algorithm demonstrates its competitiveness in addressing low/high-dimensional multi/many-objective optimization problems.
APPLIED SOFT COMPUTING
(2022)
Article
Mathematics
Zongliang Guo, Sikai Lin, Runze Suo, Xinming Zhang
Summary: In recent years, various data-driven evolutionary algorithms (DDEAs) have been proposed to solve time-consuming and computationally intensive optimization problems. DDEAs are divided into offline and online DDEAs, with offline DDEAs being widely studied and proven to have excellent performance. However, most offline DDEAs suffer from the drawbacks of requiring redundant surrogates, insufficient fitness evaluations, and increased runtime due to data generation methods. To overcome these problems, we propose a new DDEA with radial basis function networks as surrogates, utilizing a fast data generation algorithm based on clustering and an adaptive design for network parameters. The accuracy and stability of the proposed algorithm are demonstrated through numerical experiments and comparisons.
Article
Ecology
Guillaume Bastille-Rousseau, George Wittemyer
Summary: This study proposes a framework for characterizing individual variation in space-use behavior, aiming to advance understanding of the diversity of individual behavior and its influence on population organization. By analyzing 20 years of telemetry data from African elephants, the research developed four metrics to characterize differentiation in resource selection behavior within a population. The study highlights that focusing on population average responses may not capture complex individual behavioral variations, and the developed metrics provide additional information beyond mean responses, with specialization and heterogeneity being informative.
JOURNAL OF ANIMAL ECOLOGY
(2022)